Yes, in randomized trials. The size of the effect depends on which trial. In the largest one, heavy drinking days fell 13.7 percentage points more on semaglutide than on placebo. That was over 26 weeks, in 108 people with alcohol use disorder [1]. A separate register study found alcohol-related hospital admissions ran lower on the drug, at a rate ratio of 0.55 [5]. That association faded within a year of stopping — the full discontinuation data is below. Nothing here is an approved use. No dose sold for weight loss has been tested for this.
The claim started as an anecdote. It became a pharmacovigilance signal. It has now been tested twice in randomized trials. That progression is unusual, and the literature broadly agreed with the anecdote. None of it changes what these drugs are sold for. Every seller on this site’s roundup prescribes them for weight. None prescribes for this.
The larger trial
A 26-week single-center trial randomized 108 treatment-seeking adults with alcohol use disorder and obesity. The drug was weekly semaglutide at 2.4 mg or placebo. Both arms also got cognitive behavioral therapy [1]. Heavy drinking days fell 41.1 percentage points from baseline on semaglutide and 26.4 points on placebo. The treatment difference was 13.7 points (95% CI −22.0 to −5.4, p=0.0015). Eighty-one percent of participants completed the full intervention.
The placebo arm is the part worth sitting with. It fell by more than a quarter on therapy alone. Semaglutide added to that; it did not replace it. The trial cannot say what the drug would do without therapy, because it never tested that.
The smaller, earlier trial
A phase 2 trial published a year earlier took a different population. It randomized 48 adults with alcohol use disorder who were not seeking treatment. It gave nine weeks of low-dose semaglutide, topping out at 1.0 mg, and measured drinking in a laboratory task[2]. It found medium to large effects on grams of alcohol consumed (β −0.48, 95% CI −0.85 to −0.11). Peak breath alcohol concentration also fell (β −0.46, 95% CI −0.87 to −0.06). Drinks per drinking day and weekly craving also fell.
It also found nothing on two outcomes. Average drinks per calendar day did not move. Neither did the number of drinking days. People drank on the same days, and drank less when they did. That is a specific finding, not a weaker version of the big trial’s.
What the large observational data adds
A retrospective cohort of electronic health records covered 83,825 patients with obesity[4]. Semaglutide was associated with a 50% to 56% lower rate of both incidence and recurrence of alcohol use disorder over 12 months. The comparison was other anti-obesity medications. The direction was consistent across sex, age group, race, and diabetes status. It replicated in 598,803 patients with type 2 diabetes.
Cohorts of that size are good at showing a small effect is real. They are bad at showing how big it is. People prescribed one drug differ from people prescribed another, in ways no adjustment fully removes. The trials are the evidence for the size. The cohort is the evidence that it generalizes past a single clinic.
What happens after someone stops
A Swedish register study asked a question the trials above could not. What happens to the effect once a person stops taking the drug? It compared 167,026 people with type 2 diabetes against themselves, across periods on and off treatment [5]. While exposed, alcohol-related hospitalizations ran at a rate ratio of 0.55 against a person’s own unexposed periods. In the first 182 days after stopping, the association was weaker, at 0.70. By 183 to 364 days it was gone, at 1.07. That interval sits across no effect. DPP-4 inhibitors, used as a comparison, showed no such pattern. The full breakdown is in the discontinuation study.
The part that matters for anyone buying online
None of this is an approved use. The trials used brand semaglutide, at fixed doses, under supervision. The larger one added therapy. A reader who buys a compounded vial from a telehealth storefront hoping for this effect is buying a different thing. It is a preparation the FDA does not review before dispensing. The dose is often unstated. The escalation schedule is usually unpublished. We count how rarely that gets written down in who publishes a dose ladder. We list every seller that does in the published ladders tool.
Both trials also ran at doses on the escalation path, not at the starting rung. What a price does between those two points is the single most common silence in this market. It is counted in what sellers will not tell you. If the interest is appetite and reward more broadly, read the weight evidence first in what the semaglutide trials are worth.